42 ALASKAN GLACIER STUDIES
apparently the broadest area of stagnation. Here too, we find the ice front farthest from the sea, and with the broadest fringe of outwash gravels; and it is from the western margin of the lobe at its junction with the Guyot lobe, that the Yahtse River emerges, one of the largest glacial streams that pour out from the Malaspina front.
Adjoining the Agassiz lobe on the east is the Seward lobe, which, so far as our present knowledge permits us to judge, is the largest of the Malaspina lobes. Certainly no other exceeds it in size excepting possibly the Guyot, about whose extent and characteristics little is known. On the western side and front it is fringed completely by a border of ablation moraine, but on the eastern side it is not definitely separated from the Marvine lobe. It is in the Seward lobe that the Malaspina Glacier extends farthest from the mountains, reaching the ocean at Sitkagi Bluffs, but not discharging icebergs. At this point there is a low, moraine-covered cliff against whose base the waves of the Pacific Ocean break (PI. IX, A).
The easternmost lobe of the Malaspina Glacier is that dominated by the Marvine Glacier. ' It is far less pronounced than the others, and, as already stated, the boundary between it and the Seward lobe is not marked by any line of moraine. Since Seward Glacier is much larger than Marvine it might be inferred that the Seward lobe would crowd eastward to the shores of Yakutat Bay, and possibly it does so under normal conditions; but in 1906, when Marvine Glacier was advancing, the ice broken by the advance extended along the entire shore of Yakutat Bay almost to Point Manby. It is quite probable, however, that the western portion of the broken ice was really a part of the Seward lobe temporarily affected by the forward thrust of the Marvine ice. On its eastern side the Marvine lobe spreads out nearly to Yakutat Bay and to the Kwik River. It is enabled to spread out in that direction because of the absence of any obstacle; but toward the west it competes with the Seward lobe for an opportunity to spread out as a piedmont bulb. Really, therefore, the Seward and Marvine Glaciers combine to form one broad lobe, each glacier dominating one side of the lobe and at the contact so merging into one another that the line of division is not recognizable. The Marvine or eastern portion of this lobe is fringed by a narrower border of ablation moraine than common, and its front is separated from Yakutat Bay by only a narrow fringe of outwash gravel. The extreme eastern margin of the lobe, which forms the western border of the Kwik River valley, was, up to 1906, in such a stagnant condition that it was covered by a thick ablation moraine in which grew a spruce, cottonwood, and alder forest. This stagnant part represented that portion of the lobe that spread farthest east and was, therefore, least under the influence of the supply of ice which the Marvine Glacier contributes to make the lobe. The entire eastern margin, nearly to the mountain base, is covered with ablation moraine.
The Hayden Glacier, unlike the other tributaries, forms no great lobe, in this respect resembling many smaller glaciers that unite with the Malaspina. It joins the ice plateau, athwart the current of the Marvine, and almost at right angles to it. At Blossom Island the Hayden Glacier coalesces with the Marvine lobe of the Malaspina and prior to 1906 the drainage of the Blossom Island region escaped through a subglacial tunnel at this point; but the advance of Marvine Glacier in 1906 closed this tunnel and the drainage found a new outlet farther up the margin of the Hayden Glacier.
In 1905 it was not possible to say where the Marvine ice and Hayden Glacier joined, both being stagnant. In 1906 the contact between the Hayden and Malaspina Glacier